Understanding subjective urban experiences is essential for designing cities that enhance well-being. Urban design should account for the psychological effects of environments on individuals, as these significantly shape perceptions and behaviors. However, a major challenge is the limited availability of urban perception data. Recent studies have leveraged large, crowdsourced datasets like Place Pulse 2.0 (PP2) to inform machine learning (ML) models for urban perception prediction, but the accuracy and reliability of outcomes remain underexplored. There is a critical need to evaluate whether these datasets truly capture human perceptions. This study investigates the role of urban street images in understanding environmental perceptions, using the PP2 dataset and ML techniques. It explores various ML pipelines, employing TPot AutoML for model selection and 5-fold cross-validation to prevent overfitting. The goal is to identify the most efficient model that strengthens the link between automated predictions and human perception. The study also applies SHAP (SHapley Additive exPlanations) to interpret model outputs, revealing feature importance and interactions. This improves transparency and ensures ML-generated insights are actionable for urban planning. By rigorously testing ML pipelines, this research enhances predictive accuracy and contributes to the development of reliable urban design tools. The findings highlight ML’s potential in processing large-scale perception data, uncovering hidden patterns, and informing people-centered urban planning. However, further validation against real-world surveys is necessary to ensure robustness and generalizability in assessing urban perceptions.
Impact of Varying Street View Perspectives on Urban Perception: The Case of Celoria Street in Milan
Urban environments significantly influence people’s perception and walkability. Advances in computer vision and the availability of open-source Street View Imagery (SVI) have increased the use of Google Street View (GSV) for perceptual predictions and walkability assessments. However, a critical issue arises from the discrepancies between GSV images, captured from street centerlines, and SVI taken from pedestrian perspectives on sidewalks. This study examines whether people’s perceptions and street element proportions derived from GSV images align with those from sidewalk viewpoints, providing a more accurate basis for urban studies. Taking Celoria Street in Milan as a case study, two sets of 360° panoramic images were collected, one from the street center and the other from the sidewalks. These images were processed using a pre-trained perception prediction model and image segmentation techniques to generate perception responses. Dynamic Time Warping (DTW) was applied to assess the consistency between the two datasets, while Ordinary Least Squares (OLS) regression was used to analyze the impact of viewpoint changes along the street scene. Findings indicate that differences in sampling perspectives can affect urban environment assessment and perception predictions. This study highlights the potential biases of GSV data for analyzing urban environments and perceptions, advocating for more cautious use of SVI to ensure robust predictions on urban perception and walkability.
Assessing In-Motion Urban Visual Perception: Analyzing Urban Features, Design Qualities, and People’s Perception
The paper proposes a machine learning and computer vision methodology for evaluating urban visual perception during pedestrian movement. The research combines semantic image segmentation, Place Pulse 2.0 perception datasets, Google Street View imagery, and supervised machine learning models to analyze how urban physical features and spatial design qualities influence people’s perception while walking. The workflow integrates PSPNet semantic segmentation, SVM-based perception prediction, and Pearson correlation analysis to investigate relationships between urban morphology, perceived openness, imageability, enclosure, complexity, and subjective perceptions such as beauty, safety, liveliness, boredom, and depression. The study highlights how directional and panoramic visual fields produce different perceptual outcomes and demonstrates the role of AI-assisted urban analytics in understanding pedestrian experience and informing human-centered urban design.
Image Segmentation and Emotional Analysis of Virtual and Augmented Reality Urban Scenes
This paper explores how image segmentation, virtual reality, and augmented reality can be combined to evaluate citizens’ emotional responses to urban environments during participatory planning processes. The research was conducted in the Porta Romana district of Milan, an area undergoing major transformation from former industrial and railway uses into a new business-oriented neighborhood. Two complementary studies were developed: the first used VR panoramic Street View scenes shown indoors to students, while the second employed outdoor AR visualization of the proposed VITAE redevelopment project through a mobile application during a public event. Participants assessed their emotional reactions through the experiential Environmental Impact Assessment (exp-EIA) method, based on Russell’s circumplex model of emotions. At the same time, semantic image segmentation algorithms quantified the visible proportion of urban elements such as trees, buildings, roads, walls, and pavements. Results show stable correlations across both studies: trees reduce unpleasantness and emotional arousal, while buildings tend to increase unpleasantness and paved surfaces increase arousal. The study demonstrates that affordable immersive tools and AI-based scene analysis can support evidence-based urban design, citizen participation, and healthier planning strategies.
Representation Types and Visualization Modalities in Co-Design Apps
This paper’s primary goal is to analyze representation types and visualization modalities of web-based and mobile applications for collaborative processes in urban planning. To this end, a comparative study of several case studies, based on literature review, analysis of EU projects’ websites, academic/commercial websites, web-platforms, application Platforms as a Service (aPaaS), Software as a Service (SaaS) dealing with co-design for urban design and planning purposes, has been done. We analyzed 56 commercial and non-commercial apps active from 2010 to 2020 across different countries. Despite the increasing level of innovation and commercialization of Augmented/Virtual Reality solutions and immersive visualization devices in the last few years, the emerging framework of ICT solutions for participatory processes in the urban planning field is still characterized by bidimensional representations and non-immersive visualizations modalities.
How Do Nature-Based Solutions’ Color Tones Influence People’s Emotional Reaction? An Assessment via Virtual and Augmented Reality in a Participatory Process
Simulations of urban transformations are an effective tool for engaging citizens and enhancing their understanding of urban design outcomes. Citizens’ involvement can positively contribute to foster resilience for mitigating the impact of climate change. Successful integration of Nature-Based Solutions (NBS) into the urban fabric enables both the mitigation of climate hazards and positive reactions of citizens. This paper presents two case studies in a southern district of Milan (Italy), investigating the emotional reaction of citizens to existing urban greenery and designed NBS. During the events, the participants explored in Virtual Reality (VR) (n = 48) and Augmented Reality (AR) (n = 63) (i) the district in its current condition and (ii) the design project of a future transformation including NBS. The environmental exploration and the data collection took place through the exp-EIA© method, integrated into the mobile app City Sense. The correlations between the color features of the viewed landscape and the emotional reaction of participants showed that weighted saturation of green and lime colors reduced the unpleasantness both in VR and AR, while the lime pixel area (%) reduced the unpleasantness only in VR. No effects were observed on the Arousal and Sleepiness factors. The effects show high reliability between VR and AR for some of the variables. Implications of the method and the benefits for urban simulation and participatory processes are discussed.
Perception of Driving Simulations: Can the Level of Detail of Virtual Scenarios Affect the Driver’s Behavior and Emotions?
Human factors studies are becoming more and more crucial in the automotive sector due to the need to evaluate the driver.s reactions to the increasingly sophisticated driving-assistant technologies. Driving simulators allow performing this kind of study in a controlled and safe environment. However, the driving simulation.s Level of Detail (LOD) can affect the users. perception of driving scenarios and make an experimental campaign.s outcomes unreliable. This paper proposes a study investigating possible correlations between driver.s behaviors and emotions, and simulated driving scenarios. Four scenarios replicating the same real area were built with four LODs from LOD0 (only the road is drawn) to LOD3 (all buildings with real textures for facades and roofs are inserted together with items visible from the road). 32 participants drove in all the four scenarios on a fixed-base driving simulator; their performance relating to the vehicle control (i.e., speed, trajectory, brake and gas pedal use, and steering wheel), their physiological data (electrodermal activity, and eye movements), their subjective perceptions, opinions and emotional state were measured. The results showed that drivers. behavior changes in a very complex way. Geometrical features of the route and environmental elements constrain much more driving behavior than LOD does Emotions are not affected by LODs. Generally, different signals showed different correlations with the LOD level, suggesting that future studies should consider their measures while modeling the virtual scenario. It is hypothesized that scenario realism is more relevant during leisurely environmental interaction, whilst simulator fidelity is crucial in task-driven interactions.
Visual post-occupancy evaluation of a restorative garden using virtual reality photography: Restoration, emotions, and behavior in older and younger people
Natural environments have a restorative effect from mental/attentional fatigue, prevent stress, and help to revitalize psychological and physical resources. These benefits are crucial for promoting active aging, which is particularly relevant given the phenomenon of population aging in recent decades. To be considered restorative, green spaces have to meet specific requirements in ecological and psychological terms that can be assessed through Post-Occupancy Evaluation (POE), a multimethod approach commonly used by environmental psychologists and landscape architects after construction to evaluate the design outcomes from the users’ perspective. Generally, POEs consist of surveys and/or interviews accompanied by more or less structured observations of onsite users’ behavior. Despite this, various practical constraints can prevent physical access to the renovated area (e.g., weather conditions, time/resources limits, health issues, bureaucratic constraints). Exploiting digital tools for such an assessment can be a crucial support in such circumstances. The current study presents the visual POE of a restorative garden for older adults in Milan, Italy. We developed a web application, that includes the exp-EIA© patented method, which allows participants to virtually explore a visual simulation of the environment and provide their feedback. We identified 3 representative viewpoints in the redeveloped garden differing from each other for the functions and the design principles that inspired the transformation. For each point of view, we created 360° Virtual Reality photographs, that can be navigated by looking around, i.e., panning, from the standing point of each view. In connection to each virtual scene, a survey was conducted (N = 321). The focus was the psychological experience related to each viewpoint, assessed with two psychometric scales investigating the constructs of emotions (pleasure and arousal) and restoration (fascination, being away, coherence, scope, and environmental preference); such information is integrated with behavioral aspects, including the main activities prefigured by participants and their visual exploration of the VR photography. The results of the virtual exploration show that the garden is perceived as restorative, with a more intense effect in a spot purposely designed. The emotions experienced in the garden are positive and a mild level of arousal is observed. The behavioral dimension is characterized by predominantly contemplative activities and contact with nature. A cartographic representation of the psychological and behavioral data is developed, to support the maintenance of the garden.